Fault Diagnosis Method for Power Plant Generator Set Based on HPCA-PPA

Aiming at the large scale of the thermal power plant generator set, complex equipment system and different variables in each system, this paper proposes a process monitoring and fault diagnosis method based on Hierarchical Principal Component Analysis-Principal Polynomial Analysis (HPCA-PPA ) . This method not only uses the PPA method to compensate for the shortcomings of the PCA method which can not deal with the nonlinear problem well, but also uses the improved hierarchical block method to adopt different modeling methods for the sub- blocks with different characteristics. In this paper, the basic principle of the method is first clarified, followed by the method steps for process monitoring and fault diagnosis, and finally the simulation experiment. The method is applied to online monitoring and fault diagnosis of generator sets, and the correctness of the method is verified.